Phenotypic variation is a hallmark of craniofacial birth defects, but understanding the relationship between variable morphology and disease remains elusive. In this work we propose to test a model that may explain phenotypic variation within similar genotypes. This model is based on our preliminary data showing a nonlinear relationship between Sonic hedgehog (SHH) signaling and continuous phenotypic variation in the face. We hypothesize that small changes in SHH pathway activity produce large phenotypic changes that have increased variance due to heterogeneous cellular responses to compromised pathway activity. In the first Specific Aim of this grant we will focus on examining the extent to which nonlinear SHH signaling contributes to variation in cellular response. This will be done at the population and the individual cell level. In the Second and Third Specific Aims we will turn our attention to in vivo genetic models of graded SHH signaling. We will examine the phenotype of resulting embryos and we will determine the variance of the phenotypes. Our model predicts that as the nonlinearity in SHH signaling increases the potential to produce large variance also increases. The experiments designed in this application will directly test this prediction and will illuminate a potentially important mechanism that destabilizes complex developmental systems. The use of quantitative analyses such as, geometric morphometrics coupled with quantitative cellular assays and quantitative PCR, throughout this grant gives us the power to perform the analyses proposed in this application.